The proposed renewal is to center on the continued development of hardware and software systems for the real time analysis and characterization of clinical electroencephalograms. The objective of the next period of development is to gradually evolve an intelligent self-organizing clinical EEG analysis system. This will be done in four concurrent phases: (1) expansion of the present real time spectral analysis system (ADIEEG) to include additional forms of analysis such as Fast Walsh Transform, time domain extraction of transient activity, bispectrum and complex demodulation; (2) addition to ADIEEG of a module for the computation of clinically significant indices and feature complexes by band, area and hemisphere including normalized power; coherence based upon bins with significant amounts of auto and cross power; amount of transient phenomena such as spikes, sharp waves and paroxysma; symmetry ratios based upon power and coherence; and randomness ratio based upon variation in epoch to epoch power, coherence and spindle shape; (3) development of an analysis optimization executive to evaluate the significance of the various indices and feature complexes and their epoch to epoch and overall variation in order to decide which data can be discarded and to schedule those forms of analysis most likely to yield meaningful data about the distinguishing characteristics of the particular EEG record being analyzed; (4) development of a clinical EEG data base, using derived indices and feature complexes, to characterize in a general way differences between age and broad classes of clinical conditions such as the epilepsies, brain tumors, degenerative diseases and the effects of drugs. The data base will be used in real time by the analysis optimization executive to help guide its decisions. Once the system is capable of organizing its own analysis paradigms to extract clinically significant feature complexes, problems associated with classification into disease types should be greatly reduced. As development proceeds, a working version of the system in its latest operational state will be used daily to routinely analyze infant, child and adult, normal and abnormal electroencephalograms, and to study special problems such as the electroencephalographic effects of ECT, and temporal lobe foci.